American Journal of Health-System Pharmacy最新文献

Purpose: To assess the current extent of pharmacy students' involvement in collecting best possible medication histories (BPMHs) in different hospital settings, as well as the accuracy and financial implications of the collected BPMHs.

Summary: A scoping review methodology was conducted following best-practice Cochrane guidance with findings reported using the PRISMA Extension for Scoping Reviews. An appropriate search string was developed followed by a search across databases: PubMed, PubPharm, LIVIVO, PubMed Central, and Web of Science. All selected studies were published between 2000 and 2023 and met the predetermined inclusion criteria. After removing duplicates and independent screening of titles, abstracts, and full texts, 20 papers were retained. The highest number of original research papers originated from the US (n = 13, 72%). In these papers, the number of patients whose medication histories were collected ranged from 40 to 4,070 (mean, 504.6) and the number of pharmacy students who collected BPMHs ranged from 2 to 71 (mean, 17.8). Students obtained BPMHs alone or in a team with healthcare professionals (HCPs). Several papers described additional training for students. The information sources used were face-to-face patient interviews, data from community pharmacies, and interviews with HCPs and caregivers. Studies demonstrated that students can accurately collect BPMHs, identify unintentional discrepancies, and suggest healthcare interventions. Two studies identified notable cost savings from clinical interventions by pharmacy students.

Conclusion: Pharmacy students can accurately collect BPMHs. The results of this scoping review can inform the development of pharmacy curricula to enhance students' competencies and student pharmacy services that can contribute to patients' safety.

Purpose: This primer aims to serve as a foundational resource on artificial intelligence (AI) for pharmacists practicing in the emergency department (ED).

Summary: Artificial intelligence (AI) is increasingly recognized for its potential to transform healthcare, including emergency medicine (EM) and pharmacy practice. AI applications in EM include diagnostic evaluation, risk stratification, resource optimization, and therapeutic decision-making. AI's role in improving triage, diagnostics, and resource utilization in the emergency setting is discussed along with its application in the medication-use process, from prescribing to monitoring. Despite the promise of AI, significant barriers such as factual inaccuracies, ethical concerns, and data transparency prevent the widespread clinical adoption of AI tools. Challenges such as racial bias, data privacy, model transparency, and the phenomenon of hallucinations in large language model outputs are highlighted as critical considerations. AI's future success in EM will depend on responsible integration, guided by clinicians including pharmacists, and a careful consideration of ethical issues and patient-specific values.

Conclusion: Pharmacists practicing in the ED should be familiar with AI tools and should understand the importance of their role in the development, implementation, and oversight of these tools to ensure safe, effective, and equitable patient care.

Purpose: A review of services and developments over a 40-year period will be included, reflecting significant advances in the field of pharmacy-based investigational drug services (IDS) at an 836-bed teaching hospital in Boston, MA.

Summary: The institution's IDS, established in 1980, has seen significant changes due to advancements in technology and regulatory requirements. Reliance on paper-based systems has shifted to utilization of electronic and automated systems such as the Investigation Drug Service Information System (IDSIS), electronic health records, and electronic temperature monitoring. Key updates have included barcode scanning and the implementation of an automated pharmacy carousel system. Regulatory compliance with new standards has driven changes in hazardous medication handling and compounding procedures while the scope of research has expanded to include complex treatments including gene and cell therapies. The growth in research activities and technological integration has led to an increase in both the volume and capacity of the study protocols managed by IDS.

Conclusion: The institution's IDS pharmacy has evolved significantly from its inception, reflecting broader trends in research pharmacy. Technological advancements and regulatory requirements have transformed IDS practices, leading to increased efficiency and safety in the management of investigational agents. IDS's capacity to manage a broader range of research studies and its expanded role in research highlight its vital position in advancing clinical trials.

Purpose: Many hematology-oncology pharmacists (HOPs) have already reported high burnout levels. Postgraduate year 2 oncology residency program directors (ORPDs) play an important role in combating attrition in HOPs but may be vulnerable to burnout with the high demands from clinical practice and residency program responsibilities. We surveyed ORPDs to assess what resources are provided in their role and their perceptions on well-being.

Methods: A 22-question survey was sent to ORPDs listed in the ASHP residency directory. ORPDs were asked about their program and their perceptions of workload and time requirements, and a one-time Stanford Professional Fulfillment Index (PFI) was performed to assess burnout. Descriptive statistics were used for demographics, resources, incentives, time requirements, and basic assessment of the PFI. Nonparametric measurements were used to assess correlations between program characteristics and potential impact on well-being endpoints.

Results: The survey response rate was 46.0%. Of ORPDs, 74% felt that they spent at least 5 hours weekly on ORPD responsibilities and 68% felt that this was not enough time to manage them. Further, 51% felt that the ASHP standards did not provide enough time to cover responsibilities. In total, 70% of ORPDs had considered resigning within the last 12 months. The mean (SD) PFI score was 2.6 (0.6), and 35% of ORPDs reported high professional fulfillment. The mean (SD) burnout score was 1.5 (0.7), and 47% of ORPDs reported high burnout.

Conclusion: There are high levels of burnout among ORPDs and a high risk of attrition. Organizational support to assist ORPDs is essential.

Purpose: Automated dispensing cabinet (ADC) medication overrides can contribute to increased risks of drug diversion, medication errors, and waste. To reduce ADC overrides, a multidisciplinary process improvement initiative was conducted over 3 years to identify, evaluate, and monitor medication overrides, with an initial goal of quarterly 5% reductions in the override rate.

Summary: Lean Six Sigma process improvement methodology identified the root causes of inappropriate medication overrides. Through a series of interventions, both targeted and institution wide, the process improvement initiative addressed technological, process, and cultural root causes. The only clinical units excluded in this project were intraoperative areas. Targeted interventions included automated pharmacy dispensing of high-use as-needed medications and correction of interface errors between the ADC and electronic health record. System-wide interventions included updating ADC override reasons to align with policy, implementation of an approved medication override list, education, data transparency, and linking ADC override pulls to the medication administration record. The rate of overrides decreased from 6.18% at baseline to 4.41% during the initial phase of targeted interventions (29% reduction from baseline; P < 0.001), with continued improvements following organization-wide interventions to achieve an override rate of 2.13% by the control phase (65% reduction from baseline; P < 0.001). No preventable adverse drug events related to initiative changes were reported during the study period.

Conclusion: Through utilization of Lean Six Sigma methodology and involvement of a multidisciplinary process improvement team, the initiative achieved a significant and sustained reduction in the rate of medication overrides.

Purpose: Autoverification (AV) is the process in which a medication is automatically verified in the electronic health record (EHR), bypassing a pharmacist's approval. If concerns of safety and efficacy for AV are addressed, broad implementation can allow AV to be a powerful tool within a hospital system to verify high-volume, low-risk medication orders. This study aims to identify parameters for risk stratification of medications and develop a replicable framework model for identifying medications appropriate for AV at UNC Health.

Methods: The modified Delphi methodology was utilized to reach consensus on parameters used in a risk stratification tool for medication orders. This tool was applied retroactively to a sample of medication orders at UNC Health during a 1-month period (October 2023) to determine risk of adverse event for potentially autoverified orders.

Results: Fifty-five criteria met consensus for consideration for use for an AV risk appraisal tool. Results from a consensus meeting for criteria that would be used in the autoverification risk appraisal tool (AVRAT) to flag medication orders as "high-risk for AV" were age, estimated glomerular filtration rate, hemoglobin level, platelet count, body weight, and EHR documentation of continuous renal replacement therapy. Twenty medications were selected for an initial proof-of-concept evaluation of the AVRAT. Using AVRAT criteria, it was determined that a total of 6.89% of all October medication orders at UNC Health posed a low risk of a potential adverse event with AV.

Conclusion: A proof-of-concept study for the utilization of AV was effectively developed. The study results indicated that AV can possibly reduce time for medication order review across a hospital system, with a relatively small number of orders being potentially eligible for AV.

Objective: Medication errors are a worldwide public health issue. Reducing inappropriate medication use is a daily challenge for clinical pharmacists. Computerization of the medication process and the rise of artificial intelligence make it possible to develop tools to detect inappropriate prescriptions. Our main goal was to compare the performance of two machine learning models capable of predicting the probability of a prescription requiring pharmaceutical intervention (PI) using hospital data.

Methods: The study was conducted in a single hospital, with data collected over 4 years, including 2,059,847 prescription lines (a patient's entire medication regimen consists of multiple prescription lines) associated with 260,611 PIs. Two tree-based binary classification machine learning models were tested: the Light Gradient Boosting Machine (LGBM) model and the Random Forest (RF) model. The dataset was split (70% for training and 30% for testing), and training and testing were performed on the global dataset and on data stratified by medical care department.

Results: For the global dataset, the LGBM model outperformed the RF model in most metrics: accuracy (86% vs 85%), precision (80% vs 42%), specificity (97% vs 89%), area under the curve (83% vs 71%) and F1-score (58% vs 47%). However, the RF model had superior recall (53% vs 46%). Furthermore, the LGBM model trained on the global database was generally more effective than models trained on the care departments' databases.

Conclusion: The LGBM model showed superior performance in detecting inappropriate prescriptions, potentially improving the thoroughness and efficiency of prescription review. While further studies are needed to confirm these findings, the model holds significant promise for advancing hospital clinical pharmacy and enhancing patient care through optimized prescription management.

Purpose: To evaluate the impact of a population health intervention to reduce therapeutic inertia and improve hypertension control in a multipractice primary care setting.

Methods: This retrospective cohort study compares clinical and process outcomes in an intervention cohort versus a parallel comparator cohort of patients in nonintervention clinics. Centralized outreach coordinators identified patients with systolic blood pressure (BP) of >150 mm Hg, called each patient, scheduled a hypertension-focused visit with a primary care physician (PCP), then forwarded a message to the clinic-based pharmacist, who reviewed the patient record and documented clinical recommendations for hypertension control prior to the patient visit.

Results: Outreach was performed for 426 intervention patients from July to December 2022, and outcomes were compared to those in 587 usual-care patients. A higher percentage of intervention patients attended a hypertension-focused clinic visit with their PCP (57.3% vs 38.8%, adjusted P < 0.001), had hypertensive therapy addressed at that visit when their BP was above 140/90 mm Hg (63.3% vs 44.2%, adjusted P = 0.010), and achieved a BP of <140/90 mm Hg (27.9% vs 16.9%, adjusted P < 0.001) within 6 months of outreach.

Conclusion: A clinic-based population health approach reduced therapeutic inertia and improved BP control in a cohort of in hypertensive patients compared with a similar cohort of patients in clinics who did not receive the intervention.